WO2019200242A1 - Évaluation de ciment de tubage à l'aide d'une détection automatisée d'arrivées d'ondes de compression d'accrochage (p) - Google Patents

Évaluation de ciment de tubage à l'aide d'une détection automatisée d'arrivées d'ondes de compression d'accrochage (p) Download PDF

Info

Publication number
WO2019200242A1
WO2019200242A1 PCT/US2019/027195 US2019027195W WO2019200242A1 WO 2019200242 A1 WO2019200242 A1 WO 2019200242A1 US 2019027195 W US2019027195 W US 2019027195W WO 2019200242 A1 WO2019200242 A1 WO 2019200242A1
Authority
WO
WIPO (PCT)
Prior art keywords
clinging
training
cement
wellbore
machine learning
Prior art date
Application number
PCT/US2019/027195
Other languages
English (en)
Inventor
Vasileios-Marios Gkortsas
Sandip Bose
Smaine Zeroug
Original Assignee
Schlumberger Technology Corporation
Schlumberger Canada Limited
Services Petroliers Schlumberger
Schlumberger Technology B.V.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Schlumberger Technology Corporation, Schlumberger Canada Limited, Services Petroliers Schlumberger, Schlumberger Technology B.V. filed Critical Schlumberger Technology Corporation
Priority to US17/046,967 priority Critical patent/US11220897B2/en
Priority to GB2016868.8A priority patent/GB2587137B/en
Publication of WO2019200242A1 publication Critical patent/WO2019200242A1/fr
Priority to NO20201104A priority patent/NO20201104A1/en

Links

Classifications

    • E—FIXED CONSTRUCTIONS
    • E21—EARTH OR ROCK DRILLING; MINING
    • E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B47/00—Survey of boreholes or wells
    • E21B47/005—Monitoring or checking of cementation quality or level
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
    • G01V1/40—Seismology; Seismic or acoustic prospecting or detecting specially adapted for well-logging
    • G01V1/44—Seismology; Seismic or acoustic prospecting or detecting specially adapted for well-logging using generators and receivers in the same well
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01V—GEOPHYSICS; GRAVITATIONAL MEASUREMENTS; DETECTING MASSES OR OBJECTS; TAGS
    • G01V1/00—Seismology; Seismic or acoustic prospecting or detecting
    • G01V1/40—Seismology; Seismic or acoustic prospecting or detecting specially adapted for well-logging
    • G01V1/44—Seismology; Seismic or acoustic prospecting or detecting specially adapted for well-logging using generators and receivers in the same well
    • G01V1/48—Processing data
    • G01V1/50—Analysing data
    • E—FIXED CONSTRUCTIONS
    • E21—EARTH OR ROCK DRILLING; MINING
    • E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B2200/00—Special features related to earth drilling for obtaining oil, gas or water
    • E21B2200/20—Computer models or simulations, e.g. for reservoirs under production, drill bits
    • E—FIXED CONSTRUCTIONS
    • E21—EARTH OR ROCK DRILLING; MINING
    • E21B—EARTH OR ROCK DRILLING; OBTAINING OIL, GAS, WATER, SOLUBLE OR MELTABLE MATERIALS OR A SLURRY OF MINERALS FROM WELLS
    • E21B2200/00—Special features related to earth drilling for obtaining oil, gas or water
    • E21B2200/22—Fuzzy logic, artificial intelligence, neural networks or the like

Definitions

  • the subject disclosure relates to well integrity. More specifically, the subject disclosure relates to evaluation of cement in a wellbore annulus.
  • Cement evaluation in cased oil and gas wells is conducted by lowering an ultrasonic tool that implements imaging through steel casing based on the casing (quasi-Lamb) modes including a thickness- mode-dominated measurement through a pulse-echo modality, and a flexural-mode-dominated measurement through a pitch-catch modality.
  • the service is important to ascertain well integrity before proceeding to putting the well into production.
  • the measurement data is processed to extract information about the properties of the annular fill behind casing: either cement, mud, mud-contaminated cement, or gas.
  • the pulse-echo resonance data is inverted for a measure of the annular impedance, while the dispersive flexural mode data yields a modal attenuation of the casing -propagating signal across two receivers. Both measurements are combined and interpreted in terms of a solid versus liquid or gas annular fill.
  • the annular fill is comprised of a solid with a compressional wave velocity that intersects the dispersive flexural mode phase velocity curve within the signal frequency bandwidth, phase matching to a headwave in the annulus occurs and an additional contribution that follows closely the arrival from within the casing is observed: a feature referred to as a “clinging P” (with“P” referring to P- or compressional waves).
  • This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used as an aid in limiting the scope of the claimed subject matter.
  • This subject disclosure relates to a method for evaluating cement surrounding a casing in a particular wellbore.
  • the method includes obtaining flexural wave training data from various wellbores having cement surrounding casings; collecting the flexural wave training data to obtain training density displays; training a first machine learning processor to identify clinging Ps based on the training density displays; using at least one wellbore tool in the wellbore to obtain flexural wave test data; processing the flexural wave test data to obtain a density display for the wellbore; applying the density display for the wellbore to the first machine learning processor to locate a potential clinging P for the wellbore; and identifying a potential contaminated cement location and/or top of cement location for the cement of the wellbore based on the location of the potential clinging P.
  • Fig. 1 depicts a plot of a flexural wavepacket attenuation of cements as a function of acoustic impedance
  • Fig. 2 depicts flexural waveforms with clinging P manifestation obtained over a plurality of azimuthal scans of a downhole tool
  • Figs. 3A and 3B respectively depict a log showing a third-interface echo (TIE), and an impedance map as a function of azimuth manifesting a galaxy pattern which results from a TIE;
  • TIE third-interface echo
  • Fig. 4 depicts a graph of flexural wave velocity versus frequency dispersion
  • Fig. 5 depicts a time-frequency plot for near and far receivers of a flexural wave imager (FWI) measurement in the presence of clinging P (left panels) and absence of clinging P (right panels);
  • Fig. 6 depicts machine learning based detection of clinging P from FWI waveforms as a function of depth and azimuth
  • Fig. 7 depicts a training set for a convolutional neural network (CNN) which is augmented with azimuthal shifts and flips of each image from left to right and up down;
  • CNN convolutional neural network
  • Fig. 8 shows a CNN used to identify galaxy patterns and including two convolutional layers, with each layer having two leamable filters with a receptive field of 3 by 3 in order to capture a fringe;
  • Fig. 9 depicts application of the trained CNN to field data for galaxy pattern identification
  • Fig. 10 depicts an impedance map (left panel) and a resulting galaxy pattern identification (right panel) obtained by applying a sliding window in the test set;
  • Fig. 11A depicts a machine learning based workflow for the automatic detection of clinging P arrivals and for the presence of TIEs in eccentered casing configurations
  • Fig. 11B depicts a method for identifying contaminated cement from ultrasonic data obtained from a downhole tool and using trained machine learning elements that identify clinging P arrivals and the presence of TIEs.
  • analysis of a solid (cement) behind a casing is conducted by a machine learning (ML) based workflow that automates the detection of clinging P arrivals in flexural wave data to suggest the presence of contaminated cement and to identify the top of cement.
  • ML machine learning
  • feature extraction is performed on time frequency energy maps that are labelled, and a support vector machine algorithm is trained to detect candidates for clinging P arrivals.
  • the combination of the two supervised ML schemes yields an automated workflow that highlights axial and azimuthal regions where the annular fill behind casing is confirmed to be a solid with additionally a tight range for its compressional wave velocity - therefore rendering the diagnosis of the tool measurement more quantitative and more autonomous than delivered with existing inversion schemes.
  • flexural wave data is utilized for detecting clinging P arrivals, while ultrasonic pulse echo data are utilized for finding TIEs through the identification of galaxy patterns on impedance maps.
  • an Isolation Scanner tool may be used to obtain the data.
  • the Isolation Scanner tool is an ultrasonic cement evaluation tool, which combines an ultrasonic imager tool with a flexural wave imager (FWI).
  • the USI uses a pulse-echo technique from which the acoustic impedance of the cement can be determined.
  • the FWI has a pitch-catch configuration and induces a flexural Lamb mode in the casing with a transmitter and measures the reflected signals at two receivers. Combining the FWI signal attenuation calculated between the two receivers with the pulse-echo measurement and comparing it with a laboratory-based database enables the production of an interpretable image of the material in the annulus.
  • Fig. 1 provides a plot of flexural attenuation versus acoustic impedance for a well-bonded cement.
  • Z c which corresponds to the critical wave speed of the cement
  • the flexural attenuation increases linearly with the acoustic impedance of the annulus
  • the flexural attenuation drops rapidly.
  • the flexural attenuation distinguishes liquid from light cement due to the change in the slope below Z c
  • the pulse echo technique acoustic impedance
  • VDL variable density display log
  • the present disclosure relates to automating the identification of the occurrence of the clinging P feature without the intervention of an expert.
  • Current deterministic workflows are not geared to detect and exploit the presence of a clinging P for interpretation because of its sporadic nature.
  • the following disclosure presents a machine learning based workflow for the automatic detection of clinging P arrivals.
  • the presence of a clinging P in the FWI waveforms indicates that the attenuation dispersion (or attenuation as a function of frequency) is likely to exhibit a discontinuity that is associated with the intersection of the cement wave velocity with the casing flexural phase velocity dispersion curve.
  • a third-interface echo can also follow the casing flexural arrival.
  • the TIE (a reflection from the cement-rock interface in the pulse-echo signal) can arrive right after the casing event: a condition that can be confused with the presence of a clinging P event.
  • the workflow set forth hereinafter considers the disambiguation of a clinging P from a TIE by relying on an independent manifestation of TIEs called galaxy patterns in the estimated pulse-echo impedance image.
  • the latter attribute is related to the use of an electrical circuit equivalent for the inversion of the cement acoustic impedance from the pulse-echo signal under the assumption of infmitely-thick cement sheath.
  • the cement thickness is minimal as occurs when the casing is highly eccentered, the presence of the TIE perturbs the estimation of the acoustic impedance over the azimuthal range of minimal cement thickness in ways that lead to the development of features known as galaxy patterns in the inverted cement impedance. See previously incorporated United States Patent No. : 5,859,811 to Miller et al.
  • FIG. 3B A galaxy pattern in an impedance map (generated using pulse-echo data) is seen in Fig. 3B.
  • the VDL log of Fig. 3A shows that for the highlighted area of Fig. 3B with the galaxy pattern, a TIE arrives right after the casing event.
  • the clinging P phenomenon may be explained as follows. The transfer of energy from the borehole fluid to the cement sheath is carried out via the flexural wave propagating axially in the casing. At every frequency, Snell’s law requires that the horizontal slowness to be the same in the borehole fluid, the steel casing and the cement. Accordingly:
  • Fig. 4 shows the dispersion of the flexural Lamb mode with the horizontal line at a phase velocity of approximately 2750 m/s representing a wave velocity in the cement often associated with the compressional bulk wave ( v P ).
  • the intersection of this horizontal line with the dispersion curve yields the transition frequency above which there is propagation (because the flexural wave velocity is larger than the wave velocity in the cement) and below which there is evanescence.
  • This interpretation applies to a plane wave, while for a pulsed beam, which can be taken to be made of a spectrum of plane waves propagating in a small angular interval, the transition in the frequency domain is not abrupt.
  • Clinging P happens for the frequencies for which v P S ⁇ Vf iex and accordingly is found in the evanescence regime where acoustic energy does not propagate in the cement sheath but remains confined to the casing-cement interface.
  • the galaxy pattern phenomenon may be explained as follows. Galaxy patterns appear as fringes on an inverted impedance map obtained from pulse echo data, which is based on the excitation of a thickness resonance of the casing. See Sinha et al,“Geophysical prospecting using sonics and ultrasonics”, Wiley Encyclopedia of Electrical and Electronics Engineering, Ed. J. G. Webster, 340-365, (1999).
  • the annulus thickness varies as a function of azimuth and the TIE’s interfere with the thickness resonance destructively or constructively for different azimuths depending on the ultrasonic travel time through the cement, resulting in a distortion of the cement impedance from its value obtained for an infmitely-thick annulus.
  • galaxy patterns can be considered as a processing artifact. Destructive interference results in a high impedance, while constructive interference in a low one.
  • clinging Ps in order to develop a machine learning based framework to identify a clinging P, a determination needs to be made as to which features characterize its presence.
  • a clinging P appears in VDL waveforms as a low frequency echo, which interferes with the casing arrival. In one embodiment, this is the feature that is used to label each sample of a training set as having a clinging P or not.
  • an extra dimension i.e. frequency
  • Windows 1 and 3 (in the far receiver plots) and 2 (in the near receiver plots) are for low frequencies, and the energy content of windows 2 and 3 behind the casing arrival will determine the presence or not of a clinging P. Comparing the energy content of windows 2 and 3 for the cases of clinging P and no clinging P, it can be seen clearly that there exists the presence of energy next to the casing for the low frequencies when clinging P is present. This energy content for low frequencies is absent when there is no clinging P. Window 1 works as a baseline for window 3 and combining its information with window 3 shows a symmetry in the energy content around the casing for the absence of clinging P.
  • window 1 At the presence of clinging P, window 1 together with window 3 show that the casing is perturbed due to the extra echo.
  • Windows 4 (near receiver) and 5 (far receiver) are for high frequencies and correspond to the TIE’s.
  • the dimension of the windows along the time axis are based on the delay of clinging P relative to the casing and are adjusted to include the energy content that discriminates between the presence or absence of clinging P.
  • the time-frequency plots are normalized relative to the maximum energy for each depth and azimuth and the energy content of all five windows is extracted into a feature vector, which is labeled as clinging P or not.
  • a support vector machine (SVM) algorithm with a linear kernel is used to train the feature vectors, which have been labeled as clinging P or not.
  • SVM is particularly useful since it is known where to expect a clinging P in the time and frequency domains and thus features that characterize it can be extracted.
  • the SVM algorithm was applied on field data and results are shown in Fig. 6 where the left-most panel shows the machine learning based detection of clinging P from FWI waveforms as a function of depth (y axis) and azimuth (x axis).
  • the next two panels i.e., the flexural attenuation for 100 and 300 kHz
  • galaxy patterns may be automatically identified using a CNN which extracts the features automatically. Accordingly, it is not necessary to create a feature that looks like a fringe.
  • the location of the galaxy pattern unlike a clinging P, is not specific, and a CNN is well adapted to locate it.
  • a training set comprises a plurality of images with dimensions, e.g., 25 pixels in depth and 40 pixels azimuthally, labeled as galaxy pattern or not.
  • Padding may be completed azimuthally (e.g., 2 pixels at each edge) to avoid the edge effect and the training set may be augmented with azimuthal shifts and flips of each image from left to right and upside down.
  • Fig. 7 shows examples of images used in the training set, with the original image at the top left, the original image shifted by nine azimuths on the top right, the original image flipped left to right on the bottom left, and the original image flipped upside down on the bottom right.
  • FIG. 8 One embodiment of a CNN for detecting galaxy patterns is shown in Fig. 8.
  • the CNN takes as input, images which are frames of impedance maps and gives as output the probability that the image has a galaxy pattern (pl) or not (pO).
  • the shown CNN includes two convolutional layers, with each one having two learnable filters with respective field of 3 by 3 pixels. The filter size was chosen to identify one fringe, which is around 25 degrees azimuthally.
  • Each convolutional layer is followed by a rectified linear unit (ReLU), which brings the nonlinear properties to the network.
  • ReLU rectified linear unit
  • the convolutional layers followed by ReLU were used at the front end, because multiple stacked convolutional layers can develop more complex features of the input image before the destructive pooling operation. As seen in Fig.
  • the next layer is a maxpooling layer which serves as a downsampling layer, followed by a dropout layer which drops a random set of activations (e.g., 50%).
  • the representation is flattened (flatten layer) and becomes a vector followed by a fully connected (dense) layer with ReLU activation and a dropout layer.
  • a fully connected layer (dense layer with softmax activation) is provided with two values, one for each class. These values are in the real numbers space and the softmax function maps them to the probability space giving the probabilities that the input image has a galaxy pattern or not.
  • FIG. 9 an impedance map generated from pulse-echo data is shown in the left panel, and the results of a galaxy pattern identification conducted by a trained CNN is shown in the right panel.
  • the panels of Fig. 9 show that the CNN detects the galaxy pattern as a function of depth correctly.
  • galaxy patterns occur for specific azimuths. According to one embodiment, it may be desirable to localize the galaxy patterns by azimuth. Without having to retrain again, a sliding window may be used in the test set, with the sliding window covering 13 azimuths, and each pixel outside taking the average value of the whole image.
  • the window that is chosen to include the galaxy pattern is the one with the highest difference between the scores that the two classes (presence of galaxy pattern or absence of it) take. This difference is the input to the soft-max layer of Fig. 8, which maps from real numbers space to probability space. The result of the galaxy pattern localization is shown in Fig. 10 demonstrating that the CNN predicts its location correctly.
  • a machine learning based workflow for the automatic detection of clinging P arrivals and for the presence of TIEs in eccentered casing configurations may be described as follows.
  • field data obtained by, e.g., an Isolation Scanner tool is gathered.
  • the field data may be data obtained from selected cased boreholes, or more widely from many cased boreholes.
  • flexural wave data is collected into a VDL whereas at 120, the pulse echo data is processed to obtain an impedance map using an algorithm such as T3 (which is an inversion used to derive the cement impedance from the pulse-echo measurement data. See Hayman, A.J.
  • the time frequency content of these windows is extracted in feature vectors.
  • atraining set based on the feature vectors is created, and each frame is labeled as having a clinging P, as not having a clinging P, or as ambiguous.
  • a machine leaning classification algorithm such as a support vector machine (SVM) is trained on the training set to obtain trained model weights.
  • SVM support vector machine
  • a training comprising images e.g., size 25 in depth by 40 pixels in azimuth
  • can be augmented with azimuthal shifts and flips as discussed above with reference to Fig. 7).
  • a deep neural network such as a CNN is trained to identify depth sections containing galaxy patterns (as discussed above with reference to Figs. 8 and 9).
  • the galaxy patterns indicate the presence of TIEs close to the casing and may be used to disqualify a preliminary determination by the SVM that a clinging P is present.
  • a data set may be analyzed to find the presence of a clinging P as follows.
  • flexural wave data and pulse echo data for a wellbore of interest are obtained by, e.g., an Isolation Scanner tool or other tools.
  • the flexural wave data is collected into a VDL for the wellbore of interest, and at 220 the pulse echo data is processed to obtain an impedance map of the wellbore of interest.
  • the casing arrival is tracked on each frame (depth and azimuth) of the VDL and at 225, the time frequency for each frame in the VDL is computed using, e.g., a continuous wavelet transform.
  • the system conducts feature extraction for each depth and azimuth. In the time- frequency plot, windows are defined in time and frequency domain relative to the casing arrival that characterizes the presence or absence of a clinging P, and the time frequency content of these window are extracted in feature vectors. Then, at 235, a test set is created based on the feature vectors.
  • the test set is input into the trained SVM (i.e., the trained SVM is applied to the test set) to obtain a map which predicts the presence (or not) of a clinging P as a function of depth and azimuth.
  • the test data impedance map is provided as input to a previously trained CNN (i.e., the trained CNN is applied on the impedance map) to identify one or more depth sections containing a galaxy pattern, if present.
  • the galaxy pattern identifies the presence of a TIE close to the casing for a particular casing depth.
  • a sliding window approach may be used on the pulse-echo data test set to identify not only the depths, but the azimuths of the TIEs, as previously discussed with reference to Fig. 10.
  • the clinging P map is compared to the galaxy pattern map to disambiguate a clinging P identified by the SVM from a TIE.
  • a determination is made that contaminated cement is present, and a top of cement depth determination may be made.
  • processors can be performed by a processor.
  • the generation of a VDL and the generation of impedance maps will be performed by processors.
  • the SVM and CNN may be implemented by apparatus or machines such as processors.
  • the comparison of the clinging P map with the galaxy pattern map in order to identify the top of cement location may be conducted by a processor.
  • the term“processor” should not be construed to limit the embodiments disclosed herein to any particular device type or system.
  • the processor may include a computer system.
  • the computer system may also include a computer processor (e.g., a microprocessor, microcontroller, digital signal processor, or general-purpose computer) for executing any of the methods and processes described above.
  • the computer system may further include a memory such as a semiconductor memory device (e.g., a RAM, ROM, PROM, EEPROM, or Flash-Programmable RAM), a magnetic memory device (e.g., a diskette or fixed disk), an optical memory device (e.g., a CD-ROM), a PC card (e.g., PCMCIA card), or other memory device.
  • a semiconductor memory device e.g., a RAM, ROM, PROM, EEPROM, or Flash-Programmable RAM
  • a magnetic memory device e.g., a diskette or fixed disk
  • an optical memory device e.g., a CD-ROM
  • PC card e.g., PCMCIA card
  • the computer program logic may be embodied in various forms, including a source code form or a computer executable form.
  • Source code may include a series of computer program instructions in a variety of programming languages (e.g., an object code, an assembly language, or a high-level language such as C, C++, or JAVA).
  • Such computer instructions can be stored in a non- transitory computer readable medium (e.g., memory) and executed by the computer processor.
  • the computer instructions may be distributed in any form as a removable storage medium with accompanying printed or electronic documentation (e.g., shrink wrapped software), preloaded with a computer system (e.g., on system ROM or fixed disk), or distributed from a server or electronic bulletin board over a communication system (e.g., the Internet or World Wide Web).
  • a removable storage medium with accompanying printed or electronic documentation (e.g., shrink wrapped software), preloaded with a computer system (e.g., on system ROM or fixed disk), or distributed from a server or electronic bulletin board over a communication system (e.g., the Internet or World Wide Web).
  • a communication system e.g., the Internet or World Wide Web
  • the processor may include discrete electronic components coupled to a printed circuit board, integrated circuitry (e.g., Application Specific Integrated Circuits (ASIC)), and/or programmable logic devices (e.g., a Field Programmable Gate Arrays (FPGA)). Any of the methods and processes described above can be implemented using such logic devices.
  • ASIC Application Specific Integrated Circuits
  • FPGA Field Programmable Gate Arrays

Landscapes

  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geology (AREA)
  • Mining & Mineral Resources (AREA)
  • General Life Sciences & Earth Sciences (AREA)
  • Environmental & Geological Engineering (AREA)
  • Geophysics (AREA)
  • Acoustics & Sound (AREA)
  • Quality & Reliability (AREA)
  • Geochemistry & Mineralogy (AREA)
  • Fluid Mechanics (AREA)
  • Remote Sensing (AREA)
  • General Physics & Mathematics (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)
  • Investigating Strength Of Materials By Application Of Mechanical Stress (AREA)
  • Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)

Abstract

L'invention concerne l'évaluation de ciment dans un puits de forage à l'aide d'outils d'évaluation de ciment pour obtenir des informations d'image ultrasonore et des informations d'image d'onde de flexion. Des formes d'onde d'imageur à ondes de flexion sont appliquées à un système d'apprentissage machine entraîné qui prédit la présence ou l'absence de celles-ci d'un accrochage (P) en fonction de la profondeur et de l'azimut, tandis que les cartes d'impédance d'imageur ultrasonore sont appliquées à un système d'apprentissage machine entraîné qui identifie des motifs de galaxie pour des profondeurs et des azimuts, indiquant la présence d'un troisième écho d'interface (TIE) proche du tubage. Des emplacements d'accrochage (Ps) sont comparés à des emplacements de motifs de galaxie pour identifier lorsqu'un TIE pourrait provoquer une détermination d'accrochage (P) erronée. Lorsqu'un accrochage sans ambiguïté (P) est constaté, un ciment contaminé est identifié qui peut également localiser la partie supérieure du ciment dans l'espace annulaire.
PCT/US2019/027195 2018-04-12 2019-04-12 Évaluation de ciment de tubage à l'aide d'une détection automatisée d'arrivées d'ondes de compression d'accrochage (p) WO2019200242A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
US17/046,967 US11220897B2 (en) 2018-04-12 2019-04-12 Evaluating casing cement using automated detection of clinging compression wave (P) arrivals
GB2016868.8A GB2587137B (en) 2018-04-12 2019-04-12 Evaluating casing cement using automated detection of clinging compression wave (P) arrivals
NO20201104A NO20201104A1 (en) 2018-04-12 2020-10-12 Evaluating casing cement using automated detection of clinging compression wave (p) arrivals

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201862656662P 2018-04-12 2018-04-12
US62/656,662 2018-04-12

Publications (1)

Publication Number Publication Date
WO2019200242A1 true WO2019200242A1 (fr) 2019-10-17

Family

ID=68164554

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2019/027195 WO2019200242A1 (fr) 2018-04-12 2019-04-12 Évaluation de ciment de tubage à l'aide d'une détection automatisée d'arrivées d'ondes de compression d'accrochage (p)

Country Status (5)

Country Link
US (1) US11220897B2 (fr)
GB (1) GB2587137B (fr)
NO (1) NO20201104A1 (fr)
SA (1) SA520420334B1 (fr)
WO (1) WO2019200242A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111175818A (zh) * 2020-01-07 2020-05-19 中国矿业大学(北京) Co2气驱前缘位置的判断方法及其模型训练方法、装置
CN114740532A (zh) * 2022-03-18 2022-07-12 电子科技大学 一种水泥声阻抗、泥浆声阻抗的计算方法

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2592974A (en) * 2020-03-12 2021-09-15 Equanostic As Method for differentiating materials on a remote side of a partition based on the attenuation of the ultrasonic extensional zero mode
US11808136B2 (en) * 2021-05-27 2023-11-07 Halliburton Energy Services, Inc. Through-tubing, cased-hole sealed material evaluation using acoustic measurements
US20220414522A1 (en) * 2021-06-29 2022-12-29 Landmark Graphics Corporation Machine learning based ranking of hydrocarbon prospects for field exploration

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140052376A1 (en) * 2012-08-15 2014-02-20 Pingjun Guo Method for Cement Evaluation with Acoustic and Nuclear Density Logs
US20150198732A1 (en) * 2014-01-16 2015-07-16 Schlumberger Technology Corporation Cement acoustic properties from ultrasonic signal amplitude dispersions in cased wells
US20150218930A1 (en) * 2014-01-31 2015-08-06 Schlumberger Technology Corporation Method to Estimate Cement Acoustic Wave Speeds from Data Acquired by A Cased Hole Ultrasonic Cement Evaluation Tool
WO2016187242A1 (fr) * 2015-05-18 2016-11-24 Schlumberger Technology Corporation Procédé d'analyse de l'intégrité du ciment dans des colonnes de tubage par apprentissage automatique
US20170176622A1 (en) * 2015-12-18 2017-06-22 Schlumberger Technology Corporation Techniques for removing interface noise from acoustic log data

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5859811A (en) 1996-02-29 1999-01-12 Schlumberger Technology Corporation Method of analyzing waveforms
GB2593112B (en) * 2016-03-03 2021-12-01 Schlumberger Technology Bv A method of investigating a multi-string wellbore

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20140052376A1 (en) * 2012-08-15 2014-02-20 Pingjun Guo Method for Cement Evaluation with Acoustic and Nuclear Density Logs
US20150198732A1 (en) * 2014-01-16 2015-07-16 Schlumberger Technology Corporation Cement acoustic properties from ultrasonic signal amplitude dispersions in cased wells
US20150218930A1 (en) * 2014-01-31 2015-08-06 Schlumberger Technology Corporation Method to Estimate Cement Acoustic Wave Speeds from Data Acquired by A Cased Hole Ultrasonic Cement Evaluation Tool
WO2016187242A1 (fr) * 2015-05-18 2016-11-24 Schlumberger Technology Corporation Procédé d'analyse de l'intégrité du ciment dans des colonnes de tubage par apprentissage automatique
US20170176622A1 (en) * 2015-12-18 2017-06-22 Schlumberger Technology Corporation Techniques for removing interface noise from acoustic log data

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111175818A (zh) * 2020-01-07 2020-05-19 中国矿业大学(北京) Co2气驱前缘位置的判断方法及其模型训练方法、装置
CN111175818B (zh) * 2020-01-07 2020-11-27 中国矿业大学(北京) Co2气驱前缘位置的判断方法及其模型训练方法、装置
CN114740532A (zh) * 2022-03-18 2022-07-12 电子科技大学 一种水泥声阻抗、泥浆声阻抗的计算方法
CN114740532B (zh) * 2022-03-18 2023-05-30 电子科技大学 一种水泥声阻抗、泥浆声阻抗的计算方法

Also Published As

Publication number Publication date
GB2587137A (en) 2021-03-17
US11220897B2 (en) 2022-01-11
SA520420334B1 (ar) 2023-02-19
NO20201104A1 (en) 2020-10-12
GB2587137B (en) 2022-05-11
GB202016868D0 (en) 2020-12-09
US20210032973A1 (en) 2021-02-04

Similar Documents

Publication Publication Date Title
US11220897B2 (en) Evaluating casing cement using automated detection of clinging compression wave (P) arrivals
US10858933B2 (en) Method for analyzing cement integrity in casing strings using machine learning
Wang et al. Successful leveraging of image processing and machine learning in seismic structural interpretation: A review
Wu et al. Automatic fault interpretation with optimal surface voting
US10995606B2 (en) Well integrity analysis using sonic measurements over depth interval
US11378710B2 (en) Fault detection based on seismic data interpretation
US9322807B2 (en) Ultrasonic signal time-frequency decomposition for borehole evaluation or pipeline inspection
WO2017007745A1 (fr) Mesure et traitement pour détecter des couches interfaciales faibles dans des formations stratifiées contenant des hydrocarbures avec des dispositifs de diagraphie acoustique
WO2019084219A1 (fr) Procédés d'analyse de l'intégrité du ciment dans des espaces annulaires d'un puits multitubé à l'aide d'un apprentissage automatique
US11428835B2 (en) Facilitating hydrocarbon exploration and extraction by applying a machine-learning model to seismic data
EP3631524B1 (fr) Système et procédé de prédiction de joint de protection contre les défauts à partir de données sismiques
US10761064B2 (en) Detection and characterization of thin weak bedding planes in hydrocarbon-bearing laminated formations
Di et al. Semi‐automatic fault/fracture interpretation based on seismic geometry analysis
AU2020249884A1 (en) System and method for property estimation from seismic data
CN106526678A (zh) 一种反射声波测井的波场分离方法及装置
Bose et al. Acoustic evaluation of annulus b barriers through tubing for plug and abandonment job planning
EP3211594A1 (fr) Système de modélisation sismique fournissant des retouches de données d'étude sismique sur la base des comparaisons des limites de la région suspecte et procédés associés
Gkortsas et al. Machine learning for the automated detection of diagnosis-revealing features on leaky flexural wave imager data
US20110046884A1 (en) Seismic attributes for structural analysis
CA3106006C (fr) Facilitation de l`exploration de l`extraction d`hydrocarbures en appliquant un modele d`apprentissage automatique a des donnees sismiques
Zhu et al. Paleokarst caves recognition from seismic response simulation to convolutional neural network detection
Keretchashvili et al. Machine Learning-Enabled Joint Interpretation of Dipole Sonic and Borehole Image Data
US20230039040A1 (en) Method for Determining if a Wellbore Consists of Micro Annulus, Free Pipe or Solid Bonding Between the Wellbore and a Casing
Palo et al. An optimized patch-point based approach for seismic fault interpretation using CNN
US20180364382A1 (en) Similarity Determination based on a Coherence Function

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19785851

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 202016868

Country of ref document: GB

Kind code of ref document: A

Free format text: PCT FILING DATE = 20190412

122 Ep: pct application non-entry in european phase

Ref document number: 19785851

Country of ref document: EP

Kind code of ref document: A1